Review

The future of adjuvant therapy for renal cell carcinoma

Sarah J Welsh1, Tobias Janowitz1 & Tim Eisen*1

Practice Points

„„ There is a need for effective adjuvant treatment in renal cell carcinoma (RCC). Significant proportions of patients remain at high risk of recurrence after surgery, particularly those with stage III disease. Two main scoring systems are available for predicting progression after surgery: the UCLA integrated scoring system and the Mayo Clinic stage, size, grade and necrosis nomogram (also known as the Leibovich score or modified stage, size, grade and necrosis score).

„„ Trials of nontargeted agents have not shown clinical benefit within the adjuvant setting. This includes treatment with IFN‑a, IL‑2 and tumor-targeted vaccines.

„„ Several large-scale, multicenter, randomized trials are ongoing in adjuvant treatment of RCC. These include trials with the receptor tyrosine kinase inhibitors sunitinib, sora­fenib and pazopanib, in addition to the mTOR inhibitor everolimus. Successful modifications have been made to the trial protocols to take account of higher than expected drop-out rates, mainly caused by treatment-related toxicity.

„„ Improvements are needed in clinical trial design to ensure agents are more quickly and cost-effectively tested in the adjuvant setting. Many potential agents and combinations of different agents (immunomodulators, antiangiogenic agents and other signaling inhibitors) remain to be tested in the clinical setting in RCC. Multiarm, multistage clinical trial designs offer design strategies for more efficient trials of adjuvant therapy in RCC. A reliable biomarker of RCC activity would be a useful surrogate of benefit to accelerate progress in clinical trials.

1Department of Oncology, Cambridge University Health Partners, Box 193, Addenbrooke’s Hospital, Hills Road, Cambridge CB2 0QQ, UK *Author for correspondence: Tel.: +44 1223 769312; [email protected] part of

10.2217/CPR.12.39 © 2012 Future Medicine Ltd Clin. Pract. (2012) 9(4), 451–462 ISSN 2044-9038 451 Review | Welsh, Janowitz & Eisen

Summary Twenty to thirty percent of patients with stage I–III renal cell carcinoma will relapse within 5 years of surgery. Recent advances in our understanding of the molecular pathogenesis of renal cell carcinoma have led to several large randomized clinical trials investigating the role of molecularly targeted agents in the adjuvant setting. However, there are higher than expected drop-out rates due to the intolerability of side effects compared with treatment given in the metastatic setting. Additionally, significant challenges remain in the area of clinical trial design and the need to assess multiple potential therapies in a time- and cost-efficient manner, and to identify which patients are likely to benefit from adjuvant therapy.

Renal cell carcinoma (RCC) is the tenth most Patient selection is therefore an essential con- common cancer worldwide and its incidence sideration when discussing the potential role of has been rising steadily [1]. In the EU, 63,300 adjuvant therapy in RCC. It is clear that patients new diagnoses and 26,400 deaths were reported with a lower risk of relapse need to be spared in 2006 and the incidence of RCC doubled in from unnecessary treatments, while those with the period from 1975 to 2005 [2]. This may, at a higher risk of recurrence may be more likely to least in part, be due to increased rates of renal benefit from adjuvant therapy. Several prognos- tumors being discovered incidentally on radio- tic, retrospectively validated nomograms exist logic imaging performed for an unrelated reason that assess relapse risk in patients with resected [3]. The increase in medical imaging over the RCC: the UCLA integrated scoring system last decades has also allowed for the discovery (UISS) and the Mayo Clinic size, grade and of early-stage RCC in patients who are asymp- necrosis nomogram (also known as the Mayo tomatic. However, this does not fully explain score, Leibovich score or the modified stage, the increases seen for RCC overall, as although size, grade and necrosis score) [7]. These scores the greatest increase in incidence has been are particularly useful to aid appropriate patient seen for localized tumors, there have also been selection and to help to identify those patients increases in more advanced tumors [3]. This has most likely to benefit from adjuvant treatment. also contributed towards rising rates of mortal- The UISS stratification is based on the 1997 ity [2]. Rising rates of obesity and hypertension tumor, nodes and metastases score, Fuhrman (established risk factors for RCC) have been grade and Eastern Cooperative Oncology postulated to contribute. The upward smoking Group performance status and allows the clas- prevalence in earlier decades might also have sification of patients into low-, intermediate- contributed to the continuing increases in RCC, and high-risk groups for developing recurrence particularly at older ages [3]. or metastases after treatment of localized or Surgery is the standard treatment for early- locally advanced RCC [8,9]. This score includes stage RCC. If feasible, nephron-sparing surgery all histologic subtypes of RCC. It is limited with optional regional lymph node dissection by the complexity and technical demands of is the procedure of choice, or either an open obtaining molecular markers from all patients. or laparoscopic nephrectomy [4,5]. Radical sur- The UISS is currently recommended by the gery can be curative. In a recent ana­lysis of the National Comprehensive Cancer Network to National Cancer Data Base (USA), 50.6% of assist in identifying patients for adjuvant trials. patients had stage I RCC, 26.7% stages II and The Leibovich score is calculated using an III and 22.7% had stage IV RCC at presenta- algorithm that includes tumor stage, regional tion [6]. While the prognosis for stage I RCC lymph node status, tumor size, nuclear grade and is excellent, the risk of relapse in patients with histologic tumor necrosis to predict metastases- stages II and III is high, with 20–30% of all free survival in patients postnephrectomy with patients with stage I–III RCC experiencing clinically localized clear cell RCC [10,11]. The relapse after surgical excision [6]. The median authors originally included 1801 patients with a time to relapse after surgery is 1–2 years with mean follow-up of 9.7 years and showed statisti- the majority occurring within 3 years after cally significant associations between calculated i­nitial diagnosis. score and progression to metastatic RCC. The

452 Clin. Pract. (2012) 9(4) future science group The future of adjuvant therapy for renal cell carcinoma | Review

Leibovich score predicts that 10‑year survival Trials of ‘nontargeted’ agents in the is 92.5, 64.3 and 23.6% for patients with low adjuvant setting (scores of 0, 1 and 2), medium (scores of 3–5) In the last 30 years, only a few drugs have shown or high (scores of 6 or above) risk, respectively, some activity against advanced renal cancer. of developing metastases after nephrectomy. Figure 1 shows drugs licensed for treatment of This highlights the need to investigate poten- metastatic RCC. Immunomodulators such as tial therapies to reduce the risk of metastases, IFN‑a and IL-2 were initially used to control particularly in medium- and high-risk patients. metastatic RCC, stabilizing the disease for several A potential limitation of the score is its reli- years or occasionally curing it completely [13,14]. ance on histologic tumor necrosis, which does Such exceptional (although relatively infrequent) not have a standardized definition, consensus results initiated trials of immuno­modulators in for reporting or availability at many centers. It the adjuvant setting. A modest benefit in sur- should be noted that the Leibovich score is also vival was reported with IFN-a and with IL-2 applicable only to patients with clear cell RCC. therapy in the context of metastatic RCC, but The Leibovich score was shown to be slightly these immune modulators do not currently have superior to the UISS nomogram in one valida- a defined role in the adjuvant setting. In rand- tion study of 388 patients with an accuracy of omized trials, adjuvant IFN‑a and recombinant 0.830 compared with 0.760 [12]. IFN-a2b have been shown to not contribute to There is currently no evidence from survival or relapse-free survival [15,16]. For exam- p­roperly designed randomized controlled tri- ple, a Phase III trial investigating adjuvant IL-2 als to s­upport the use of adjuvant treatment in in high-dose bolus form was closed early because l­ocalized RCC. However, several large rand- an interim ana­lysis revealed that disease-free sur- omized clinical trials are currently investigating vival (DFS) was not affected [17]. In a subsequent the role of new targeted agents in the adjuvant study, adjuvant IL-2 in low-dose subcutaneous setting after s­uccesses in treating metastatic form was also shown to be ineffective with respect RCC. This article aims to briefly review past to prolonging DFS [18]. Whether the lack of effec- and present trials of adjuvant treatment in RCC tiveness of immunomodulators in the adjuvant and s­uggest what can be learnt from these expe- setting was due to the low objective response rate riences for the future of adjuvant treatment (even in the metastatic setting), patients being in RCC. included despite being at low risk of recurrence,

Immunotherapies Antiangiogenic agents

Temsirolimus mTOR inhibitor IFN-α and bevacizumab Everolimus mTOR inhibitor Sorafenib and sunitinib Pazopanib IL-2 and IFN-α High-dose IL-2 Multitargeted TKIs Cytokine activity US FDA approved based Multitargeted reported on Phase II data TKI Bevacizumab VEGF antagonist

1980 1985 1990 1995 2000 2005 2010

VHL gene isolated

Figure 1. Timeline for drugs licensed for the treatment of metastatic renal cell carcinoma. Immunotherapies were licensed in the 1980s and 1990s. In recent years, targeted therapies aimed at angiogenesis have been approved for clinical use. TKI: Tyrosine kinase inhibitor; VHL: von Hippel–Lindau.

future science group www.futuremedicine.com 453 Review | Welsh, Janowitz & Eisen

or whether insufficient follow-up was performed strategies for the treatment of metastatic RCC. to see an effect, is unclear. Central to the biology of sporadic RCC is loss A more gentle method of immuno­modulation of function of the von Hippel–Lindau (VHL) and its application to adjuvant therapy in tumor suppressor gene, located on chromo- RCC has come in the form of tumor vaccines. some 3p (reviewed in [24]). VHL was identified Galligioni et al. investigated the use of autolo- in patients with familial VHL disease, an auto- gous irradiated tumor cells mixed with bacillus somal dominant cancer syndrome associated Calmette–Guèrin as an adjuvant strategy and with the development of a number of tumors found no statistically significant improvement in including conventional RCC [25]. Recent com- overall survival (OS) or DFS [19], but a multi- prehensive genetic studies suggest very high rates center, Phase III, randomized controlled trial of (>95%) of VHL involvement through mutation, adjuvant autologous tumor cell vaccine conducted methylation and loss of heterozygosity ana­lysis, in Germany showed a statistically significant DFS such that VHL loss of function might provide benefit[20] . In the latter investigation, 379 patients a molecular basis for classification as clear cell with pT2–3b pN0–3 M0 disease were included in RCC [26,27]. The VHL gene product functions the ana­lysis and 5‑year progression-free survival in the hypoxia inducible factor (HIF) pathway, (PFS) was 77.4 and 67.8% in the vaccine and the forming a multiprotein complex that principally control groups, respectively. That study, however, functions by ubiquitinating HIF-a leading to its has been widely criticized because 174 patients proteasomal degradation [28,29]. Since the expres- were lost to follow-up after randomization and sion of the HIF pathway is primarily controlled also because differences in OS were not analyzed. by levels of HIF-a subunits, loss or inactivation Another vaccine strategy has focused on the of VHL leads to high levels of HIF-a subunits, use of heat shock proteins. The heat shock pro- and increased activity of the HIF pathway [27–31]. tein peptide complex HSPPC-96 (vitespen) was HIF is a transcription factor that co-ordinates developed from autologous tumors in RCC. the response of cells to low oxygen levels and is Following encouraging results in a Phase II critical for tumor cell survival (reviewed in [32]). In trial, a 728‑patient, multicenter, open-label, response to low oxygen levels, HIF stimulates up randomized Phase III trial compared adjuvant to 100 genes, which increase cellular metabolism HSPPC-96 with observation following nephrec- and cell survival pathways, and activate angio­ tomy and found no difference in recurrence-free genesis (new blood vessel formation) to improve survival after a median follow-up of 1.9 years [21]. oxygenation and nourishment. This includes The use of hormonal therapy has also been VEGF, which results in RCCs being characterized explored as a potential adjuvant treatment of as highly vascular tumors. high-risk RCC. In a prospective randomized Therapies targeting this pathway (so-called study of 136 patients, medroxyprogesterone ‘targeted’ therapies) have proven a highly effective acetate was found to provide no benefit with therapeutic strategy in RCC, improving the out- regard to disease recurrence and was associated look for patients with advanced disease (reviewed with significant toxicity [22]. in [33]). Inhibition of the VEGF pathway has been Therefore, currently no adjuvant therapies achieved via monoclonal antibodies targeted to have shown clear evidence of DFS improvement bind VEGF (bevacizumab) or through intra­ [16,23] and active surveillance has remained the cellular inhibition of VEGF signaling through the standard of care after partial or radical nephrec- use of small molecule tyrosine kinase inhibitors tomy. However, these trials demonstrated that (TKIs) that target the intracellular kinase domains large, multicenter trials are feasible and that of the VEGF receptors (VEGFR1–3), such as well-powered trials that look for statistical differ- sora­fenib, sunitinib, pazopanib and axitinib [34– ences in DFS and OS are possible in the adjuvant 40]. Receptor tyrosine kinases are essential for the setting in renal cancer. transduction of extracellular signals into the cell. Studies have shown improvement of PFS and OS Rationale for investigating targeted agents resulting in regulatory approval to treat patients in the adjuvant setting in RCC with metastatic RCC with these drugs. Recent advances in our understanding of the Levels of HIF-a (and VEGF) have also been molecular pathogenesis of RCC have led to shown to be increased in RCC via activation of the the successful development of new therapeutic mTOR–PI3K pathway [40–42]. Agents targeting

454 Clin. Pract. (2012) 9(4) future science group The future of adjuvant therapy for renal cell carcinoma | Review

mTOR, such as temsirolimus and everolimus, ƒƒ The PROTECT trial is a double-blind, rand- also exert antiangiogenesis effects and have been omized, multicenter study comparing treat- approved to treat patients with metastatic RCC ment with pazopanib for 1 year versus placebo [43,44]. In patients who have progressed on first- [105]. The study will include 1500 patients with line therapy with a VEGF pathway antagonist, localized or locally advanced RCC after everolimus has been shown to offer clinical ben- nephrectomy. The primary end point is DFS, efit (modest prolongation of PFS compared with with OS, safety, health outcome and quality of placebo in a randomized Phase III study) [44,45]. life as secondary end points; Overall, targeted drugs have a relatively ƒƒ The ARISER trial (completed ana­lysis is favorable toxicity profile in the advanced dis- expected) studied cG250 (WX-G250; an ease setting and are usually orally bioavailable immunoglobulin G1 antibody targeting car- (reviewed in [33]). Therefore, encouraged by bonic anhydrase IX) in a multicenter, rand- positive outcomes in the metastatic setting, tar- omized, placebo-controlled, Phase III trial in geted therapies are being tested in the adjuvant clear cell RCC patients following complete or setting. Six large-scale, randomized trials are in partial surgical removal of the affected kidney, progress (Table 1): in patients with no detectable metastases [106]. ƒƒ The S-TRAC trial is a double-blind, rand- Carbonic anhydrase IX is a HIF downstream omized, multicenter study comparing sunitinib target gene that is expressed in all clear cell with placebo in 720 patients at high risk of RCC, but is not detected in normal kidney or relapse (based on UISS criteria) after nephrec- most other normal tissues [46]. It may be tomy [101]. The primary end point is DFS and involved in cell proliferation and transformation. the study will be completed in 2015; Results from the above adjuvant trials will also ƒƒ The SORCE trial is a double-blind, rand- be important to answer concerns emerging from omized, multicenter study with three treat- laboratory reports that manipulation of the HIF ment arms comparing treatment with sora­ and VEGF pathways either genetically or using fenib for 1 year versus treatment with sora­fenib angio­genesis inhibitors may potentially accelerate for 3 years versus treatment with placebo [102]. metastasis [47]. Studies using pancreatic neuro­ The study will include 1656 patients with endocrine and glioblastoma multiforme tumor resected RCC with intermediate and high risk cells have shown that resistance to angiogenesis of developing metastatic disease, based on the inhibitors may occur via a process of evasive resist- Leibovich score (score of 3–11). The primary ance. Resistant clones develop using alternative end point is DFS and recruitment to the study proangiogenic pathways such as the FGF pathway is expected to be completed in 2013; and have been shown to be more invasive and demonstrate increased dissemination compared ƒƒ The ASSURE trial is a double-blind, random­ with tumor cells not treated with angiogenesis ized, multicenter study comparing treatment inhibitors [47]. This is therefore a concern for adju- with placebo versus sorafenib or sunitinib [103]. vant studies if treatment resulted in recurrence of The study will include 1923 patients with disease with tumor cells d­emonstrating a more pT1b–4 tumors or with fully resected node- aggressive phenotype. positive disease. Patients will be stratified into high- and very-high-risk groups. The primary end point is DFS, with OS and translational What lessons have been learned so far? Studies carried out to date have shown that the studies as secondary end points. This trial is use of targeted therapies in the adjuvant setting in the process of reporting; presents a different set of challenges from their use ƒƒ The SWOG-S0931 trial is a double-blind, ran- in the metastatic setting. It is well known that the domized, multicenter study comparing treat- TKIs are associated with a number of common ment with everolimus for 1 year versus placebo toxicities. Common to all currently used TKIs is [104]. The study will include 1218 patients with that they are multitargeted agents, inhibiting a high- and very-high-risk groups after radical number of receptor kinases, including PDGFR‑a or partial nephrectomy. The primary end point and ‑b, stem cell factor receptor (KIT), RET and is DFS, with OS, toxicity and translational FMS-like tyrosine kinase-3 (Flt‑3), in addition to studies as secondary end points; VEGF receptors, with varying potency [48]. This

future science group www.futuremedicine.com 455 Review | Welsh, Janowitz & Eisen [101] Ref. [105] [103] [102] [108] [106] [104] Outcome measures tolerability QoL Enrollment (n) 1218 OS, DFS, toxicity 1923 OS, DFS, toxicity 1656 OS, DFS, toxicity 7201500 DFS, OS, DFS, OS, toxicity, 140 DFS, OS 864 OS, DFS, QoL Registration/ completion 05/2010 08/2013 05/2006 04/2016 05/2007 2013 09/2006 11/2015 10/2010 10/2015 12/2009 11/2013 07/2004 09/2013 Multicenter, doubleMulticenter, blind, randomized, placebo controlled doubleMulticenter, blind, randomized, placebo controlled blind, randomized, placebo controlled doubleMulticenter, blind, randomized, placebo controlled blind, randomized, placebo controlled open-Multicenter, label, randomized Multicenter, doubleMulticenter, blind, randomized, placebo controlled NIH; NCI, ECOG, SWOG, Cancer and Leukemia Group B, NCIC GroupClinical Trials Sciences; Universities: Beijing, Central South, Sichuan, China Medical, Tianjin Medical, Shanghai Zhejiang,Jiao Tong, Fudan, Huazhong; Jiangsu Province Centers for Disease Control and Prevention; Peking Union Medical College Hospital; Changhai Hospital Sorafenib or Sunitinib Girentuximab Industry; Wilex ); DFS: Disease-free DFS: ); survival; ECOG: Eastern Cooperative Oncology Group; NCI: National Cancer Institute; NCIC: National Cancer Institute of Canada; Overall OS: survival; recruiting recruiting Southwest Oncology Group.

Acronym Status Intervention Funding body and sponsors Design Table 1. Details 1. of clinicalTable trials of targeted agents in adjuvant treatment of renal cell carcinoma. CTD trial number NCT01120249 SWOG-S0931 RecruitingNCT00326898 Everolimus ASSURE NIH; SWOG, NCI NCT00492258 Active, not SORCENCT00375674 Recruiting S-TRAC SorafenibNCT01235962 Recruiting PROTECT Medical Research Council UK SunitinibNCT01041482 Recruiting AGuo double Multicenter, Pazopanib Industry; Pfizer Industry; GlaxoSmithKline Recruiting Sorafenib double Multicenter, Chinese Academy of Medical CTD: Clinical Database Trials (see ClinicalTrial website [107] QoL: Quality of life; SWOG: NCT00087022 ARISER Active, not

456 Clin. Pract. (2012) 9(4) future science group The future of adjuvant therapy for renal cell carcinoma | Review

lack of specificity brings with it a number of com- compared with the metastatic setting, it is dif- mon side effects, often termed ‘off-target’ effects, ficult to predict whether targeted agents will be including hypothyroidism, hand–foot syndrome, similarly effective in patients with no evidence diarrhea, stomatitis and anorexia. Others, such of remaining disease after surgery. Therefore, as hypertension and lethargy, may in fact repre- patient selection is an important issue. To date, sent ‘on-target’ toxicities. In the metastatic set- the studies in the adjuvant setting are focused on ting, many patients require dose reductions (or patients with intermediate- or high-risk disease, stop therapy altogether), which has been shown based mainly on factors such as stage, size and to negatively impact on both quality of life and grade. Currently, no biomarker is available that survival [49]. predicts the effectiveness of angio­genesis inhibi- Unsurprisingly, toxicities have been shown tors in RCC patients. However, several mark- to be particularly relevant to use of TKIs in ers are being developed to tailor antiangio­genic the adjuvant setting. However, early data from therapy, although none have been validated in adjuvant trials have demonstrated that the toxic- large, prospective trials [50,51]. Such a biomar- ity profile may be different than that observed ker would be very helpful to identify and select during treatment in the metastatic setting with those patients most likely to benefit from adju- an often higher drop-out rate than expected but vant treatments but also to spare patients poten- lower event rate than in standard practice [Eisen T, tially distressing toxicities when they would be Pers. Comm.]. For example, toxicities in the SORCE unlikely to benefit. trial have been higher than predicted with 35% Likewise, relevant biologic end points of of patients experiencing significant toxicities over therapy and progression remain elusive in RCC the whole course of the trial, compared with a treatment. Therefore, an important aspect of predicted rate of 20% [102]. Importantly, such tox- all adjuvant trials is the correlative science that icities have caused patients to drop out of adjuvant attempts to identify such markers. Yuasa et al. trials with the majority of drop-outs occurring have recently reviewed the role of biomarkers within the first 12 weeks of treatment. Both the in predicting the response to sunitinib treat- SORCE and ASSURE trials have had to increase ment in the metastatic setting [52]. In addition enrollment targets to account for these drop-out to clinical factors, as defined by the Memorial rates (e.g., from 1332 to 1923 in the case of the Sloane–Kettering Cancer Center (MSKCC) ASSURE trial). score, genetic factors affecting the pharmaco- Interestingly, long-term toxicity has been dynamics of sunitinib (e.g., the activity of drug less of a problem and actions have been taken efflux pumps and metabolizing enzymes), in to respond to early toxicity concerns in the first addition to the baseline activity of the targeted 3 months: both the SORCE and ASSURE tri- pathway within each patient, have been shown als have introduced half-dose lead-in periods to affect the response to sunitinib. Selection of with an emphasized dose-modification schedule appropriate biomarkers is therefore critical, espe- with early prophylactic measures advised (see the cially within the adjuvant setting. The follow- ClinicalTrial website [107] for details). The aim ing factors have been included in the adjuvant of the first 12 weeks has also been modified such trials described above: functional status of the that the focus during this period is to identify tumor vasculature and markers of tissue ang- a dose that the patient can tolerate well, rather iogenesis and apoptosis and associated mark- than achieving the maximum dose. There has ers (e.g., immuno­histochemistry for microves- also been a realization that dose interruption sel density and apoptosis); pharmacokinetics early is helpful to ensuring that patients remain and genotyping (measurement of circulating on treatment. This has since had a positive impact proangiogenic markers and cells, drug lev- on drop-out rates [Eisen T, Pers. Comm.]. Since adju- els, CYP3A4/5, B-RAF and VEGF polymor- vant studies take a considerable length of time phisms); and proteomics on plasma and serum to run, given the low event rate in this setting, and hypermethylation markers (P16, VHL and it is especially important to respond to concerns others in urine and tumor tissue). The results and address issues quickly to ensure results are will be essential to enable appropriate patient meaningful and relevant. selection and to guide therapy with the aim Equally, since the role of angiogenesis is less of personalizing treatment to each individ- well understood in the earlier stages of RCC ual patient. The ability to enrich trials for an

future science group www.futuremedicine.com 457 Review | Welsh, Janowitz & Eisen

appropriate treatment population would also the question of whether further risk stratifica- reduce cost and likely save time, particularly in tion is needed, with high-risk patients receiving the adjuvant setting. early postoperative therapy whereas medium‑risk patients may be able to delay treatment. How can clinical trials of adjuvant The traditional approach to clinical trial treatment in RCC be improved? design is to test each agent one by one in sepa- Despite the availability of multiple treatment rate controlled trials. An alternative is the multi- options, many questions and challenges remain arm multistage (MAMS) trial design whereby for clinical trials of adjuvant treatment in RCC. several novel treatments are compared [53]. The In effect, these include optimal sequencing of two approaches are compared in Figure 2. In the available agents (which is the best first-line the MAMS trial design, single-agent, single- or subsequent therapy for a given patient) in arm Phase II trials are followed by a single addition to how to design trials with appropri- MAMS trial of all combination therapies. For ate comparison arms and end points, and ensure example, using current standard trial protocols, that well-tolerated and effective drug combina- new agents (T1, T2 and T3) would be tested tions are identified (including those in different initially in three separate single-agent, single- classes and with different modes of action, such arm Phase II trials, followed by three single- as targeted agents, cytotoxic agents and immu- arm combination Phase II trials. The MAMS notherapy). Ongoing trials may answer some of design rolls the Phase II assessment of the activ- these questions but many agents and combina- ity of combination therapy into the same trial tions remain untested and are awaiting trials. as the Phase III assessment of effectiveness. Additionally, the current trials do not address The MAMS model therefore would require 1300 patients compared with 2100 patients with the traditional model (a saving of 800 patients). Phase Traditional trial Analyses MAMS trial MAMS trials therefore require fewer patients design design and less overall time, since patients are rand- T1 T2 T3 C T1 T2 T3 omized from the start and different agents are tested concurrently rather than sequentially. II P P P There is also a shared control group and, at interim analyses, arms may be dropped if evi- dence so far suggests that they are unlikely to C T1 P P P be effective (futility), or if sufficient evidence of effectiveness has been found already (efficacy). P P There is also no delay between Phase II and III assessments and MAMS trials require fewer P applications for finance and approvals (one pro- C T2 P S S tocol, one grant application, one clinical trial Time agreement submission [per country], one ethics III application [per country] and one research and development approval [per site]). Importantly, P the trials infrastructure remains in place for a C T3 series of clinical questions and does not have to be rebuilt each time. Therefore, MAMS trials are more efficient and cheaper. An additional advantage for adjuvant trials in RCC patients is P that more treatments can be tested with a lim- ited set of patients (e.g., in different subtypes of RCC). MAMS trials are also more popular Figure 2. Comparison of traditional and multiarm multistage clinical trial design. among patients as they stand a greater chance of MAMS trial designs require fewer patients for control arms and have multiple points being allocated to a new treatment. of analyses and can therefore be completed sooner. However, the MAMS trial design also has some C: Control/placebo; MAMS: Multiarm multistage; P: Point of primary analysis; drawbacks, including the need for co­operation S: Point of secondary analysis; T: Treatment/drug to be tested. between different commercial companies, which

458 Clin. Pract. (2012) 9(4) future science group The future of adjuvant therapy for renal cell carcinoma | Review

could cause significant delays in set-up time passed the prespecified intermediate hurdle for for trials if not addressed. For example, a trial activity. Since the possibility of stopping recruit- involving those agents currently being tested in ment to arms is an integral part of the trial’s the adjuvant setting would require cooperation design and is built into the approved protocol, between four major pharmaceutical companies this did not constitute a substantial amendment (Pfizer, Bayer, GlaxoSmithKline and Novartis). and was initiated straight away. Additionally, a This problem has recently been addressed by new arm was later added comparing abirater- the NIH’s National Center for Addressing one/prednisolone to control after abiraterone Translational Sciences (NCATS) in the develop- received marketing authorization in the USA ment of a collaborative program that will match and in the EU from September 2011 [56]. The researchers with a selection of pharmaceutical STAMPEDE trial has therefore demonstrated industry compounds to help scientists explore that the MAMS approach is a practical as well new treatments for patients. The pilot program as a theoretical advance. incorporates innovative template agreements designed to streamline the legal and admin- Conclusion istrative process for participation by multiple There is a need for adjuvant treatment in RCC organizations. These template agreements aim to reduce relapse rates. Significant Phase III trials to reduce time, cost and effort, as well as allow on the adjuvant use of targeted therapy in RCC greater participation than traditional partner- are ongoing. They investigate those therapies that ships, and may be helpful in the context of a have shown benefit in the treatment of meta- MAMS trial design. static RCC. Already, these trials have identified An additional issue that could potentially the need to respond to concerns and to address delay a MAMS trial is that the trial could not issues quickly to ensure results are meaningful start until several agents are available. In renal and relevant in this setting. Significant challenges cancer, there are currently several agents for remain, particularly in the area of clinical trial which it would be useful to conduct a head-to- design and the need to assess multiple potential head trial within the adjuvant setting, includ- therapies in a time- and cost-efficient manner, ing sunitinib/sora­fenib, pazopanib, axitinib and and to identify which patients are likely to benefit everolimus; only the ASSURE trial will answer from adjuvant therapy. We suggest the MAMS which of sunitinib or sora­fenib may be more trial design as a possible model to address these effective in the adjuvant setting. The MAMS problems. The next few years will see reports on trial design also allows for the addition of extra a number of adjuvant trials and therefore offer arms at later dates to respond quickly to new the promise of better outcome from RCC after agents becoming available. surgical treatment. This trial design has been used in the STAMPEDE trial to compare hormone therapy Future perspective alone with a combination of hormone therapy Within the next 5–10 years, the role of adjuvant and either zoledronic acid, docetaxel (or both), therapy for renal cancer will have been clari- celecoxib or abiraterone in prostate cancer fied. Important questions regarding which, if [54,55]. The trial aims to recruit 4000 men with any, of the receptor TKIs (or the mTOR inhibi- advanced prostate cancer between September tor everolimus) leads to improved outcomes for 2005 and December 2013. It has recruited well renal cancer patients will have been answered and appears to be on-track to determine which, by the ongoing large clinical trials. Importantly, if any, therapy is beneficial for use in this setting. the tolerability of these agents within the adju- Importantly, when the trial came to its second vant setting will have been assessed, which will planned intermediate activity ana­lysis (a pre- allow clinicians to guide development of future determined point at which sufficient data had therapies with improved knowledge regarding the been accrued to allow an intermediate ana­lysis level of acceptable toxicity in this patient group. to take place) the STAMPEDE Trial Steering Questions regarding scheduling of adjuvant ther- Committee decided that recruitment should be apy are also expected to be answered, including stopped to the celecoxib-containing trial arms whether treatment can be delayed in medium- due to a lack of sufficient benefit. Recruitment risk patients compared with those identified to to the other arms continued unchanged as these have higher-risk disease. Increased use of imaging

future science group www.futuremedicine.com 459 Review | Welsh, Janowitz & Eisen

and biomarkers capable of accurately assessing renal cancer remains an exciting area and the clinical response will allow clinicians to more next 10 years will see key questions answered in accurately identify those patients most likely to the field. benefit from treatment, while sparing those likely to have poor outcomes or those likely to suffer Financial & competing interests disclosure from intolerable treatment-related side effects. T Eisen has a consultancy or advisory role for Bayer, Given the likely ongoing need to quickly and Pfizer, Roche, GlaxoSmithKline (GSK) and AVEO; has cost-effectively translate potential therapies from stock ownership at AstraZeneca; honoraria from Roche, preclinical and early clinical studies into the Bayer, Pfizer, GSK and AVEO; and receives research clinical setting, assessment methods of clinical funding from Astrazeneca, GSK, Pfizer and Bayer. The efficacy will need to become more streamlined. authors have no other relevant affiliations or financial The MAMS clinical trial model is proposed by involvement with any organization or entity with a finan- the authors as a potential strategy. Collaborations cial interest in or financial conflict with the subject matter between different institutions, pharmaceuti- or materials discussed in the manuscript apart from those cal companies and academic research institu- disclosed. tions will need to evolve to achieve these goals. No writing assistance was utilized in the production of However, the future of adjuvant treatment in this manuscript.

9 Zisman A, Pantuck AJ, Wieder J et al. Risk 13 Minasian LM, Motzer RJ, Gluck L, Mazumdar References group assessment and clinical outcome M, Vlamis V, Krown SE. Interferon alfa-2a in Papers of special note have been highlighted as: algorithm to predict the natural history of advanced renal cell carcinoma: treatment n of interest patients with surgically resected renal cell results and survival in 159 patients with n n of considerable interest carcinoma. J. Clin. Oncol. 20, 4559–4566 long-term follow-up. J. Clin. Oncol. 11, 1 Jemal A, Siegel R, Ward E, Hao Y, Xu J, Thun (2002). 1368–1375 (1993).

MJ. Cancer statistics, 2009. Cancer J. Clin. n First to describe the UCLA integrated 14 Fisher RI, Rosenberg SA, Fyfe G. Long-term 59, 225–249 (2009). scoring system to predict risk of relapse after survival update for high-dose recombinant 2 Ferlay J, Autier P, Boniol M, Heanue M, surgery in renal cell carcinoma (RCC) interleukin-2 in patients with renal cell Colombet M, Boyle P. Estimates of the cancer patients. carcinoma. Cancer J. Sci. Am. 6(Suppl. 1), incidence and mortality in Europe in 2006. S55–S57 (2000). 10 Frank I, Blute ML, Cheville JC, Lohse CM, Ann. Oncol. 18(3), 581–592 (2007). Weaver AL, Zincke H. An outcome 15 Messing EM, Manola J, Wilding G et al. 3 Chow WH, Devesa SS, Warren JL, Fraumeni prediction model for patients with clear cell Phase III study of interferon alfa-NL as JF Jr. Rising incidence of renal cell cancer in renal cell carcinoma treated with radical adjuvant treatment for resectable renal cell the United States. JAMA 281, 1628–1631 nephrectomy based on tumor stage, size, carcinoma: an Eastern Cooperative Oncology (1999). grade and necrosis: the SSIGN score. J. Urol. Group/Intergroup trial. J. Clin. Oncol. 21, 1214–1222 (2003). 4 Hollingsworth JM, Miller DC, Dunn RL 168, 2395–2400 (2002). 16 Pizzocaro G, Piva L, Colavita M et al. et al. Surgical management of low-stage renal n Demonstrated that tumor stage, tumor size, cell carcinoma: technology does not supersede nuclear grade and histologic tumor necrosis Interferon adjuvant to radical nephrectomy in Robson stages II and III renal cell carcinoma: a biology. Urology 67, 1175–1180 (2006). could predict progression to metastatic RCC multicentric randomized study. J. Clin. Oncol. 5 Leibovich BC, Blute ML, Cheville JC et al. in patients with clear cell RCC. 19, 425–431 (2001). Nephron sparing surgery for appropriately 11 Leibovich BC, Blute ML, Cheville JC et al. 17 Clark JI, Atkins MB, Urba WJ et al. Adjuvant selected renal cell carcinoma between 4 and Prediction of progression after radical high-dose bolus interleukin-2 for patients with 7 cm results in outcome similar to radical nephrectomy for patients with clear cell renal high-risk renal cell carcinoma: a Cytokine nephrectomy. J. Urol. 171, 1066–1070 (2004). cell carcinoma: a stratification tool for Working Group randomized trial. J. Clin. 6 Kane CJ, Mallin K, Ritchey J, Cooperberg prospective clinical trials. Cancer 97, Oncol. 21, 3133–3140 (2003). MR, Carroll PR. Renal cell cancer stage 1663–1671 (2003). migration: ana­lysis of the National Cancer 18 Majhail NS, Wood L, Elson P, Finke J, Olencki n Forms the basis for development of the Data Base. Cancer 113, 78–83 (2008). T, Bukowski RM. Adjuvant subcutaneous Leibovich score used in some of the ongoing interleukin-2 in patients 7 DeVita VT Jr, Hellman S, Rosenberg SA. adjuvant trials. with resected renal cell carcinoma: a pilot study. Cancer Principles and Practice of Oncology (8th 12 Ficarra V, Novara G, Galfano A et al. Clin. Genitourin. Cancer 5, 50–56 (2006). Edition). Lippincott Williams and Wilkins, The ‘stage, size, grade and necrosis’ score is Philadelphia, PA, USA (2008). 19 Galligioni E, Quaia M, Merlo A et al. Adjuvant more accurate than the University of immunotherapy treatment of renal carcinoma 8 Zisman A, Pantuck AJ, Dorey F et al. California Los Angeles Integrated Staging patients with autologous tumor cells and Improved prognostication of renal cell System for predicting cancer-specific survival bacillus Calmette–Guèrin: five-year results of a carcinoma using an integrated staging system. in patients with clear cell renal cell carcinoma. prospective randomized study. Cancer 77, J. Clin. Oncol. 19, 1649–1657 (2001). BJU Int. 103, 165–170 (2009). 2560–2566 (1996).

460 Clin. Pract. (2012) 9(4) future science group The future of adjuvant therapy for renal cell carcinoma | Review

20 Jocham D, Richter A, Hoffmann L et al. 30 Krieg M, Haas R, Brauch H, Acker T, Flamme 42 Laughner E, Taghavi P, Chiles K, Mahon PC, Adjuvant autologous renal tumour cell vaccine I, Plate K. Up-regulation of hypoxia-inducible Semenza GL. HER2(neu) signaling increases and risk of tumour progression in patients with factors HIF-1a and HIF-2a under normoxic the rate of hypoxia-inducible factor 1a renal-cell carcinoma after radical nephrectomy: conditions in renal carcinoma cells by von (HIF-1a) synthesis: novel mechanism for Phase III, randomised controlled trial. Lancet Hippel–Lindau tumor suppressor gene loss of HIF-1-mediated vascular endothelial growth 363, 594–599 (2004). function. Oncogene 19, 5435–5443 (2000). factor expression. Mol. Cell Biol. 21, 21 Wood C, Srivastava P, Bukowski R et al.; 31 Turner KJ, Moore JW, Jones A et al. Expression 3995–4004 (2001). C-100–112 RCC Study Group. An adjuvant of hypoxia-inducible factors in human renal 43 Hudes G, Carducci M, Tomczak P et al. autologous therapeutic vaccine (HSPPC-96; cancer: relationship to angiogenesis and to the Temsirolimus, interferon-a, or both for vitespen) versus observation alone for patients von Hippel–Lindau gene mutation. Cancer Res. advanced renal cell carcinoma. N. Engl. J. Med. at high risk of recurrence after nephrectomy for 62, 2957–2961 (2002). 356, 2271–2281 (2007). renal cell carcinoma: 32 Semenza GL. Oxygen homeostasis. Wiley 44 Motzer RJ, Escudier B, Oudard S et al. Phase 3 a multicentre, open-label, randomised Interdiscip. Rev. Syst. Biol. Med. 2(3), 336–361 trial of everolimus for metastatic renal cell Phase III trial. Lancet 372, 145–154 (2008). (2010). carcinoma: final results and ana­lysis of 22 Pizzocaro G, Piva L, Di Fronzo G et al. 33 Kim WY, Kaelin WG Jr. Molecular pathways prognostic factors. Cancer 116, 4256–4265 Adjuvant medroxyprogesterone acetate to in renal cell carcinoma – rationale for targeted (2010). radical nephrectomy in renal cancer: 5‑year treatment. Semin. Oncol. 33(5), 588–595 45 Motzer RJ, Escudier B, Oudard S et al. results of a prospective randomized study. (2006). Efficacy of everolimus in advanced renal cell J. Urol. 138, 1379–1381 (1987). 34 Singer EA, Gupta GN, Srinivasan R. Targeted carcinoma: a double-blind, randomised, 23 Atzpodien J, Schmitt E, Gertenbach U et al. therapeutic strategies for the management of placebo-controlled Phase III trial. Lancet 372, Adjuvant treatment with interleukin-2- and renal cell carcinoma. Curr. Opin Oncol. 24(3), 449–456 (2008). interferon-alpha2a-based 284–290 (2012). 46 Shuch B, Li Z, Belldegrun AS. Carbonic chemoimmunotherapy in renal cell carcinoma anhydrase IX and renal cell carcinoma: n Useful review of current and future targeted post tumour nephrectomy: results of a prognosis, response to systemic therapy, and therapies in RCC. prospectively randomised trial of the German future vaccine strategies. BJU Int. 101, 25–30 Cooperative Renal Carcinoma 35 Pantuck AJ, Zeng G, Belldegrun AS, Figlin (2008). hemoimmunotherapy Group (DGCIN). Br. J. RA. Pathobiology, prognosis, and targeted 47 Pàez-Ribes M, Allen E, Hudock J et al. Cancer 92, 843–846 (2005). therapy for renal cell carcinoma: exploiting the Antiangiogenic therapy elicits malignant hypoxia-induced pathway. Clin. Cancer Res. 9, 24 Drucker BJ. Renal cell carcinoma: current progression of tumors to increased local 4641–4652 (2003). status and future prospects. Cancer Treat. Rev. invasion and distant metastasis. Cancer Cell 31, 536–545 (2005). 36 Escudier B, Eisen T, Stadler WM et al. 15(3), 220–231 (2009). Sorafenib in advanced clear-cell renal-cell 25 Latif F, Tory K, Gnarra J et al. Identification of 48 Bhargava P, Robinson MO. Development of carcinoma. N. Engl. J. Med. 356, 125–134 the von Hippel–Lindau disease tumor second-generation VEGFR tyrosine kinase (2007). suppressor gene. Science 260, 1317–1320 inhibitors: current status. Curr. Oncol. Rep. (1993). 37 Motzer RJ, Hutson TE, Tomczak P et al. 13(2), 103–111 (2011). Sunitinib versus interferon-a in metastatic 26 Young AC, Craven RA, Cohen D et al. 49 Ravaud A, Bello CL. Exposure–response renal-cell carcinoma. N. Engl. J. Med. 356, Analysis of VHL gene alterations and their relationships in patients with metastatic renal 115–124 (2007). relationship to clinical parameters in sporadic cell carcinoma receiving sunitinib: maintaining conventional renal cell carcinoma. Clin. Cancer 38 Escudier B, Pluzanska A, Koralewski P et al. optimum efficacy in clinical practice. Res. 15(24), 7582 (2009). Bevacizumab plus interferon-a-2a for Anticancer Drugs 22(5), 377–383 (2011). treatment of metastatic renal cell carcinoma: a 27 Maxwell PH, Wiesener MS, Chang GW et al. 50 McGuire BB, Fitzpatrick JM. Biomarkers in randomised, double-blind Phase III trial. The tumour suppressor protein VHL targets renal cell carcinoma. Curr. Opin Urol. 19, Lancet 370, 2103–2111 (2007). hypoxia-inducible factors for oxygen- 441–446 (2009). dependent proteolysis. Nature 399(6733), 39 Sternberg CN, Davis ID, Mardiak J et al. 51 Zurita AJ, Jonasch E, Wu HK, Tran HT, 271–275 (1999). Pazopanib in locally advanced or metastatic Heymach JV. Circulating biomarkers for renal cell carcinoma: results of a randomized n n Explains how levels of hypoxia inducible vascular endothelial growth factor inhibitors in Phase III trial. J. Clin. Oncol. 28, 1061–1068 factor (HIF)-a subunits are regulated by renal cell carcinoma. Cancer 115, 2346–2354 (2010). VHL and is a seminal paper in HIF biology. (2009). 40 Zhong H, Chiles K, Feldser D et al. 28 Maxwell PH, Dachs GU, Gleadle JM 52 Yuasa T, Takahashi S, Hataki K, Yonese J, Modulation of hypoxia-inducible factor 1a et al. Hypoxia-inducible factor-1 modulates Fukui I. Biomarkers to predict response to expression by the epidermal growth factor/ gene expression in solid tumors and influences sunitinib therapy and prognosis in metastatic phosphatidylinositol 3-kinase/PTEN/AKT/ both angiogenesis and tumor growth. Proc. renal cancer. Cancer Sci. 102, 1949–1957 FRAP pathway in human prostate cancer cells: Natl Acad. Sci. USA 94, 8104–8109 (1997). (2011). implications for tumor angiogenesis and 29 Wiesner M, Munchenhagen P, Berger I et al. therapeutics. Cancer Res. 60, 1541–1545 53 Parmar MK. Speeding up the evaluation of Constitutive activation of hypoxia-inducible (2000). new agents in cancer. J. Natl Cancer Inst. genes related to overexpression of hypoxia- 100(17), 1204–1214 (2008). 41 Zundel W, Schindler C, Haas-Kogan D et al. inducible factor-1a in clear cell renal Loss of PTEN facilitates HIF-1-mediated gene n Good review and explanation of the multiarm carcinomas. Cancer Res. 61, 5215–5222 (2001). expression. Genes Dev. 14, 391–396 (2000). multistage (MAMS) trial design.

future science group www.futuremedicine.com 461 Review | Welsh, Janowitz & Eisen

54 James ND, Sydes MR, Clarke NW et al. TheTreatment Of Patients At High Risk Of 105 A Study to Evaluate Pazopanib as an Adjuvant STAMPEDE: Systemic Therapy for Recurrent Renal Cell Cancer (S-TRAC). Treatment for Localized Renal Cell Advancing or Metastatic Prostate Cancer – a http://clinicaltrials.gov/ct2/show/ Carcinoma (RCC) (PROTECT). multi-arm multi-stage randomised controlled NCT00375674 http://clinicaltrials.gov/ct2/show/ trial. Clin. Oncol. 20(8), 577–581 (2008). 102 Sorafenib in Treating Patients at Risk of NCT01235962 55 James ND, Sydes MR, Clarke NW et al. Relapse After Undergoing Surgery to Remove 106 Monoclonal Antibody Therapy (Rencarex®) Systemic therapy for advancing or metastatic Kidney Cancer. in Treating Patients Who Have Undergone prostate cancer (STAMPEDE): a multi-arm, http://clinicaltrials.gov/ct2/show/ Surgery for Non-metastatic Kidney Cancer. multistage randomized controlled trial. BJU NCT00492258 http://clinicaltrials.gov/ct2/show/ Int. 103(4), 464–469 (2009). 103 Sunitinib or Sorafenib in Treating Patients NCT00087022 56 de Bono JS, Logothetis CJ, Molina A et al. With Kidney Cancer That Was Removed By 107 US NIH clinical trials database. Abiraterone and increased survival in Surgery. www.clinicaltrials.gov metastatic prostate cancer. N. Engl. J. Med. http://clinicaltrials.gov/ct2/show/ 108 A Multi-Center, Non-Controlled Clinical 364(21), 1995–2005 (2011). NCT00326898 Study of Sorafenib Adjuvant Therapy in 104 Everolimus in Treating Patients With Kidney Advanced Renal-Cell Carcinoma (AGuo). „„Websites Cancer Who Have Undergone Surgery. http://clinicaltrials.gov/ct2/show/ 101 A Clinical Trial Comparing Efficacy And http://clinicaltrials.gov/ct2/show/ NCT01041482 Safety Of Sunitinib Versus Placebo For NCT01120249

462 Clin. Pract. (2012) 9(4) future science group